Androgen receptor and osteoglycin gene expression predicting prognosis of metastatic prostate cancer

Abstract

This study aimed to identify the predictive factors associated with oncological outcomes in metastatic hormone-sensitive prostate cancer-related genes. A nomogram for predicting prostate cancer-specific survival (CSS) was constructed based on biopsy samples from 103 patients with metastatic hormone-sensitive prostate cancer. We analyzed the association between clinical data and mRNA expression levels. The nomogram was externally validated in another cohort (n = 50) using a concordance index. Based on the cutoff value, determined by a receiver operating characteristic analysis, longer CSS was observed in the high osteoglycin and androgen receptor expression level groups (> 1.133 and > 0.00; median CSS, 85.3 vs. 52.7 months, p = 0.082, and 69.1 vs. 32.1 months, p = 0.034, respectively), compared with that of the low expression level groups. The nomogram predicting CSS included hemoglobin (≥ 13.7 g/dL or < 13.7 g/dL), serum albumin (≥ 3.1 g/dL or < 3.1 g/dL), serum lactate dehydrogenase (≥ 222 IU/L or < 222 IU/L), total Japan Cancer of the Prostate Risk Assessment score, androgen receptor expression level, and osteoglycin expression level. The concordance indices for internal and external validations were 0.664 and 0.798, respectively. A nomogram that integrates expression levels of androgen receptors and osteoglycin to predict CSS in metastatic hormone-sensitive prostate cancer was established.